Hyperspectral Target Detection via Adaptive Joint Sparse Representation and Multi-Task Learning with Locality Information

被引:22
作者
Zhang, Yuxiang [1 ]
Wu, Ke [1 ]
Du, Bo [2 ]
Zhang, Liangpei [3 ]
Hu, Xiangyun [1 ]
机构
[1] China Univ Geosci, Inst Geophys & Geomat, Hubei Subsurface Multiscale Imaging Key Lab, Wuhan 430074, Peoples R China
[2] Wuhan Univ, Sch Comp, Wuhan 430079, Peoples R China
[3] Wuhan Univ, State Key Lab Informat Engn Surveying Mapping & R, Wuhan 430079, Peoples R China
来源
REMOTE SENSING | 2017年 / 9卷 / 05期
基金
中国国家自然科学基金;
关键词
hyperspectral image; target detection; multi-task learning; sparse representation; locality information; IMAGE CLASSIFICATION; FEATURE-EXTRACTION; REDUCTION; SELECTION; MODEL;
D O I
10.3390/rs9050482
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Target detection from hyperspectral images is an important problem but encounters a critical challenge of simultaneously reducing spectral redundancy and preserving the discriminative information. Recently, the joint sparse representation and multi-task learning (JSR-MTL) approach was proposed to address the challenge. However, it does not fully explore the prior class label information of the training samples and the difference between the target dictionary and background dictionary when constructing the model. Besides, there may exist estimation bias for the unknown coefficient matrix with the use of l(1)/l(2) minimization which is usually inconsistent in variable selection. To address these problems, this paper proposes an adaptive joint sparse representation and multi-task learning detector with locality information (JSRMTL-ALI). The proposed method has the following capabilities: (1) it takes full advantage of the prior class label information to construct an adaptive joint sparse representation and multi-task learning model; (2) it explores the great difference between the target dictionary and background dictionary with different regularization strategies in order to better encode the task relatedness; (3) it applies locality information by imposing an iterative weight on the coefficient matrix in order to reduce the estimation bias. Extensive experiments were carried out on three hyperspectral images, and it was found that JSRMTL-ALI generally shows a better detection performance than the other target detection methods.
引用
收藏
页数:19
相关论文
共 35 条
  • [1] [Anonymous], 2008, SIAM J OPTIM
  • [2] Ben-David S., 2003, P C COMP LEARN THEOR
  • [3] Target oriented dimensionality reduction of hyperspectral data by Kernel Fukunaga-Koontz Transform
    Binol, Hamidullah
    Ochilov, Shuhrat
    Alam, Mohammad S.
    Bal, Abdullah
    [J]. OPTICS AND LASERS IN ENGINEERING, 2017, 89 : 123 - 130
  • [4] Enhancing Sparsity by Reweighted l1 Minimization
    Candes, Emmanuel J.
    Wakin, Michael B.
    Boyd, Stephen P.
    [J]. JOURNAL OF FOURIER ANALYSIS AND APPLICATIONS, 2008, 14 (5-6) : 877 - 905
  • [5] Multitask learning
    Caruana, R
    [J]. MACHINE LEARNING, 1997, 28 (01) : 41 - 75
  • [6] Accelerated Gradient Method for Multi-Task Sparse Learning Problem
    Chen, Xi
    Pan, Weike
    Kwok, James T.
    Carbonell, Jaime G.
    [J]. 2009 9TH IEEE INTERNATIONAL CONFERENCE ON DATA MINING, 2009, : 746 - +
  • [7] Sparse Representation for Target Detection in Hyperspectral Imagery
    Chen, Yi
    Nasrabadi, Nasser M.
    Tran, Trac D.
    [J]. IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING, 2011, 5 (03) : 629 - 640
  • [8] Preprocessing EO-1 Hyperion hyperspectral data to support the application of agricultural indexes
    Datt, B
    McVicar, TR
    Van Niel, TG
    Jupp, DLB
    Pearlman, JS
    [J]. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2003, 41 (06): : 1246 - 1259
  • [9] Automated Hyperspectral Cueing for Civilian Search and Rescue
    Eismann, Michael T.
    Stocker, Alan D.
    Nasrabadi, Nasser M.
    [J]. PROCEEDINGS OF THE IEEE, 2009, 97 (06) : 1031 - 1055
  • [10] On the impact of PCA dimension reduction for hyperspectral detection of difficult targets
    Farrell, MD
    Mersereau, RM
    [J]. IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2005, 2 (02) : 192 - 195